From Micro to Global: Gauging the Health of E‑Learning Ecosystems with MOSAIC

Online learning environments can be thought of as living ecosystems – complex networks of learners, content, technologies, and institutional supports that interact and evolve. Like natural ecosystems, they have indicators of “health” and sustainability: robust knowledge flow, adaptability to change, and the ability to nurture growth over time. In fact, digital learning technologies hold the potential to disrupt and reconfigure long-standing educational structures [1], demanding new ways to understand what it means for an online learning system to thrive. How can we assess the health and sustainability of e-learning ecosystems at multiple scales, from individual lessons up to global e-campuses? In this post, we explore this question through the lens of the MOSAIC framework – a model of Modular, Outcome-based, Stackable, Adaptive, Integrated Curriculum [1] – and introduce new conceptual terminology to describe the dynamics of learning in the digital era. We propose ideas like pedagogical metabolism, digital cognitive resilience, and adaptive curricular entanglement to illuminate how e-learning ecosystems process knowledge, withstand disruptions, and weave together learning experiences across scales. Throughout, we maintain a theoretical stance inspired by contemporary learning science, while keeping a practical eye on how educators and institutions might apply these ideas via MOSAIC’s scalable design principles.

E-Learning Ecosystems as Living Systems

What does it mean to treat an online learning environment as an ecosystem? An ecosystem is a holistic entity – whether a forest or a virtual classroom – whose health depends on the interplay of its components. In education, this perspective urges us to look beyond isolated courses or technologies and see the entire web of interactions that constitute learning. At the heart of an ecological perspective on learning is the need to make connections across formal, informal, and everyday learning [3]. A healthy learning ecosystem seamlessly connects individual study sessions and class activities with broader curricula and even real-world practice. This aligns with MOSAIC’s emphasis on integrated, stackable modules that fit together “like the interlocking tiles of a mosaic” into a cohesive whole [1].

We can draw on biological metaphors to articulate key qualities of a flourishing e-learning ecosystem:

• Pedagogical metabolism – the rate and efficiency with which a learning system processes knowledge inputs into learning outcomes. Just as an organism metabolizes nutrients, an online course metabolizes content (readings, videos, lectures) and interactions into understanding and skills. A high pedagogical metabolism means information flows smoothly: students actively engage with materials, receive feedback, and apply concepts, with minimal “waste” in the form of confusion or disengagement. For example, a well-designed tutorial that rapidly adjusts to a learner’s responses and provides instant feedback is exhibiting a fast, efficient pedagogical metabolism. In contrast, a stagnant discussion forum where questions go unanswered for weeks signals a slow metabolism. Monitoring this could involve analytics on content engagement and turnaround time on feedback – essentially, checking the vital signs of the learning process.
  
• Digital cognitive resilience – the capacity of an online learning ecosystem to support deep, durable learning in the face of challenges or change. This encompasses the system’s ability to adapt when disruptions occur (such as a sudden shift to remote learning, or the introduction of a new technology) and to help learners persist through difficulties. An ecosystem with strong digital cognitive resilience provides multiple pathways and support mechanisms for learning. If one approach fails – say a student doesn’t grasp a concept from the video lecture – the system offers alternatives (additional examples, peer discussion, tutoring AI) to ensure understanding is eventually achieved. Recent global disruptions like the COVID-19 pandemic have underscored why such resilience is critical; indeed, some argue higher education should “never be the same” post-crisis, instead embracing more flexible, technology-enhanced models [2]. A resilient e-learning ecosystem treats technology not as a simple substitute for face-to-face teaching, but as an opportunity to reimagine pedagogy in ways that make learning more robust to shocks.
  
• Adaptive curricular entanglement – the intricate interconnectedness of learning modules and experiences across different levels of instruction, designed to be responsive to learner needs. In a healthy ecosystem, courses aren’t siloed; their curricula are entangled in productive ways. Skills learned in one module flow into the next, and themes recur with increasing complexity across a program. This concept mirrors the “integrated” aspect of the MOSAIC framework, ensuring that each component of the curriculum is contextually linked to others [1]. The entanglement is adaptive in that the connections can bend and reorganize in response to data – if learners consistently struggle with a concept in Course A that is needed in Course B, the system can adjust, perhaps by introducing a bridging module or reinforcing that concept through extra practice. Over time, an adaptively entangled curriculum self-optimizes, much like a neural network strengthening important connections. The result is a learning journey that feels coherent and personalized, as opposed to a disjointed series of requirements.

These theoretical notions give us a language to discuss e-learning ecosystem health. But how do they play out in practice at different scales? Next, we examine scalability – from the micro level of individual lessons, to courses and communities, to institutions, and ultimately the global stage – to see how the MOSAIC principles help maintain ecosystem vitality at each level.

Scalability from Micro to Macro (and Beyond)

One strength of the MOSAIC framework is that it inherently addresses scalability. By design, MOSAIC’s modular and stackable approach ensures that small learning units (lessons, modules) can aggregate into larger structures (courses, certificates, degrees) without losing their integrity [1]. We can thus assess e-learning ecosystem health on multiple levels of magnification:

Micro-Scale: Lessons and Modules

At the most granular level, we have individual learning activities – a video lesson, an interactive quiz, a discussion thread, a single module in an online course. This is where the pedagogical metabolism is most immediate. Key health questions at this scale include: Is each lesson achieving its intended learning outcomes (and how do we know)? Are learners engaged and receiving feedback within a timely cycle? A healthy lesson or module is clear in purpose, rich in interaction, and tightly aligned with outcomes (MOSAIC’s O for outcome-based). For instance, a micro-lesson might pose a real-world problem, guide the student through resources, then prompt them to apply what they learned, all in a short sequence. If students can successfully solve the problem and reflect on it, we know the knowledge was effectively processed – a sign of a strong pedagogical metabolism at work.

Because MOSAIC emphasizes Modular design, each micro-unit should stand on its own and also fit into broader competencies [1]. This modularity means we can stack small pieces into larger credentials – an idea increasingly evident in the rise of micro-credentials. Micro-credentials provide learners with the ability to quickly learn and implement new skills to stay current, often making learning more attainable and flexible [4]. In practice, offering a 2-week module that confers a digital badge (for example, in data visualization or instructional design) can be seen as a micro-scale ecosystem. Its health might be measured by how many learners earn the badge, how they rate the experience, and whether they continue to the next module. Studies show that these bite-sized, competency-based learning units can increase accessibility and motivation, acting as “bridges” that attract learners who might not commit upfront to a long degree program [4]. In other words, a strong micro-ecosystem feeds the larger ecosystem by broadening participation.

To sustain quality at this level, outcome-based design is crucial – every module should have clear, measurable outcomes (knowledge or skills) that tie into the program’s bigger goals. Immediate data can be gathered through embedded assessments and learning analytics: quiz scores, time on task, discussion posts. These data are the “sensors” of the micro-ecosystem. If the data indicate, for example, that 80% of students missed Question 3 in a quiz, the system flags a possible breakdown in understanding. An instructor (or an AI assistant) can then intervene to clarify or provide additional resources. This kind of responsiveness – effectively, a feedback loop – keeps the metabolism healthy and prevents small issues from accumulating. It’s worth asking: could an AI tutor within a module monitor each learner’s progress and dynamically adjust the difficulty or provide hints, acting like a personal trainer for cognitive fitness? Such adaptive technology at the lesson level would exemplify MOSAIC’s Adaptive principle, ensuring that each learner gets the support they need before moving on.

Meso-Scale: Courses and Class Communities

Zooming out, the next level is the course as a whole (or a learning community, such as a cohort in a class). Here multiple modules interconnect, and social dynamics enter the picture. A course is more than a collection of content; it’s an ecosystem of people, pedagogy, and technology interacting over an extended time. Health at this meso-scale can be seen in metrics like sustained student engagement throughout the course, the depth of discussions, peer-to-peer support, and the achievement of course-level outcomes. One could say the course has its own metabolism: how regularly are ideas being exchanged, assignments completed, feedback given and incorporated? If a course ecosystem goes “stale” (low posting activity, few logins, minimal questions asked), that’s akin to a metabolic slowdown indicating low energy and engagement.

A critical factor here is the sense of community. Learners who feel connected to each other and their instructor form a collaborative learning community, which MOSAIC identifies as a key puzzle piece for student success [1]. Research shows that sustained peer-to-peer interaction drives deeper learning and helps online students persist [1]. In practical terms, a healthy course ecosystem fosters active discussion forums, group projects, and peer feedback loops. These not only enrich understanding but also create accountability and belonging. The concept of digital cognitive resilience is very much at play in community interactions: when a learner hits a roadblock, peers or mentors in the class can help them bounce back, offer explanations, or simply share that they too struggled but overcame it. Such social support structures act like an immune system for the course, catching and addressing problems early. We might ask: how do we gauge the “social health” of an online class? Possible indicators include network analytics (are all students engaged or are some isolated?), sentiment analysis in discussion (are the interactions positive and constructive?), or even simple attendance/activity statistics over time.

Moreover, a well-designed course features adaptive curricular entanglement internally. This means the assignments, resources, and assessments are woven together so that insights from one activity inform the next. For example, an initial quiz might determine which of two project options a student is guided into, based on their demonstrated strengths or interests – entangling the path of the course with the learner’s profile. The course syllabus may not be entirely fixed on day one; there might be branching scenarios or optional modules that allow adaptation. This adaptability can be powered by instructor judgment and/or learning analytics. It aligns with MOSAIC’s adaptive and outcome-driven ethos: the course dynamically adjusts to ensure outcomes are met for diverse learners.

It’s worth noting that not everything can or should be automated or adaptive – human pedagogical judgment is vital. An instructor monitoring the class might notice a pattern (e.g., many struggled with Week 3’s concept) and decide to host an extra live review session. In essence, the instructor is acting as a gardener tending the course ecosystem, guided by data but also by experience. Healthy ecosystems at the meso-scale benefit from this blend of human and data-informed care.

Macro-Scale: Programs and Institutions

At the institutional or program level (e.g., a fully online master’s program or an entire university’s e-learning operations), we encounter a higher-order ecosystem – a network of courses, support services, faculty, and infrastructure that together deliver education. Here, sustainability often comes to the fore. A sustainable e-learning ecosystem can maintain its quality and improve over multiple cohorts; it can scale up to serve more students or scale down to personalize; it can weather external pressures (like shifting job market demands or technological disruptions) by evolving its curriculum. Key health indicators at this macro-scale might include student retention and graduation rates, learner satisfaction, post-graduation outcomes (employment or further study), and the continuous innovation of curriculum and pedagogy. Essentially, is the ecosystem thriving year after year, producing successful graduates and new knowledge, or is it deteriorating (high dropout rates, outdated content, declining enrollments)?

The MOSAIC framework offers a vision for a “living” curriculum at the program level – one that is modular and flexible enough to evolve. For instance, stackable credentials allow learners to accumulate certificates and micro-credentials en route to a full degree [1]. This stackability is a sign of health: it means the ecosystem has multiple entry and exit points, accommodating learners in different life circumstances. A student might start with a 3-course certificate, then later return to stack it into a degree – and the ecosystem supports that pathway smoothly (much like a thriving coral reef supports both small fish and larger ones at different life stages). The presence of stackable, modular curriculum components is an indicator of adaptability and learner-centered design at the institutional level.

Another pillar of a healthy institutional ecosystem is the support and mentoring infrastructure. Online learners can feel like “faces in the crowd,” so proactive advising and mentorship are crucial to sustain them. In fact, integrated mentoring is a piece of the MOSAIC puzzle: programs implementing MOSAIC pair students with dedicated mentors or advisors, embedded throughout the learning journey. This has a direct effect on health: research indicates that stronger advising and pastoral support in virtual settings significantly reduce attrition [8]. Imagine an online program where each student has an AI-enhanced academic advisor that checks in on their progress, answers common questions 24/7, and flags a human advisor when a deeper intervention is needed. That kind of support scaffold can dramatically improve the sustainability of the ecosystem by catching issues (academic or personal) that might lead a student to drop out, and addressing them in a timely manner.

From a pedagogical metabolism standpoint, an institution shows health by how effectively it turns inputs into outputs at scale. Inputs include faculty expertise, content, technology tools, and student effort. Outputs are skilled graduates, research insights, and innovation in practice. An efficient pedagogical metabolism at this level might involve robust faculty development (so instructors are well-versed in online pedagogy), rapid incorporation of feedback into course improvements, and agile governance that can approve curricular changes or new courses as needed. Some forward-looking institutions have embraced an iterative design approach: every offering of an online course is reviewed with data from learning analytics and student feedback, then refined for the next cycle. In this way, the curriculum adapts continuously, never staying static. Over a few iterations, the program “learns” and adapts just as the students do – a meta-metabolism, if you will, where the institution learns how to learn.

We should also consider the institutional culture as part of the ecosystem. Is there a mindset of innovation and support for e-learning? Are successes recognized and scaled up, while failures are seen as learning opportunities? A sustainable ecosystem requires active cultivation by leadership – setting policies that encourage open educational resources, investing in technology infrastructure, and rewarding teaching excellence in online formats. For example, if an analytics system shows that a certain course significantly improved its completion rate after adopting a new adaptive learning tool, a healthy response is to celebrate and study that success, then propagate effective practices to other courses. This echoes the MOSAIC philosophy of integrating best practices and continuous improvement across the board [1].

Mega-Scale: Global e-Campuses and Lifelong Networks

Finally, we zoom out to the global scale – the emerging ecosystem of interconnected online learning experiences worldwide. In today’s context, a learner in Illinois might be taking a MOOC from a platform in California, collaborating with peers in Europe, while also enrolled in a local community college course. The global e-campus is a reality, enabled by the internet and increasingly by open educational resources and massive online communities of practice. Here, the health of the ecosystem is about inclusivity, interoperability, and global knowledge sharing. In the age of digital technology and AI, the learning ecosystem is interconnected, employing both online and offline resources to enable learning to take place anywhere, anytime, via individualized pathways [5]. This vision captures a sustainable global learning ecosystem that breaks down barriers of geography and time, allowing lifelong learning to flourish.

One way to view the global learning ecosystem is as a network of networks. Each institution or platform is a node, and learners often cross between them (for example, bringing a Coursera certificate into a university degree, or leveraging a coding bootcamp to get job-ready). Adaptive curricular entanglement at this scale means creating links between different learning experiences across the world. Initiatives around common credit frameworks, credential recognition, and international partnerships contribute to this entanglement. A healthy global ecosystem would allow a learner to seamlessly weave their learning journey from multiple sources – formal and informal – into a coherent tapestry. In practice, this might mean global standards for micro-credentials or widespread adoption of e-portfolios that travel with the learner. We see early signs: employers accepting digital badges, universities forming consortiums to honor each other’s online courses. These are analogous to ensuring genetic diversity and cross-pollination in a biological ecosystem, preventing any single learning environment from becoming too insular.

Another indicator at the mega-scale is the accessibility and equity of learning worldwide. If certain regions or populations are left out due to the digital divide, the global ecosystem’s health is compromised (just as an ecosystem suffers if one species or resource is wiped out). Sustainability here means not only environmental sustainability (using e-learning to reduce carbon footprints, for instance) but also social sustainability: reaching underserved communities, supporting learners with disabilities via adaptive tech, and providing education that is culturally relevant across contexts. We might measure this through global enrollment figures, diversity statistics of online program participation, and the availability of multilingual and low-bandwidth educational resources.

It’s also at this scale that the pedagogical metabolism concept gains a new dimension: knowledge creation. Healthy global learning ecosystems aren’t just consuming knowledge, they’re producing it. When learners in different parts of the world can collaborate, they generate new ideas, open-source projects, research, and innovations at an unprecedented pace. Many-to-many digital networks make it possible for social learning to involve more people at greater speeds than the social learning spaces of the print era [1]. In other words, the metabolism of knowledge circulation worldwide has accelerated – ideas can propagate in days via webinars, forums, and publication platforms, whereas in the past they might stay localized for years. The challenge and opportunity here is to harness that fast metabolism for collective learning: global hackathons, international research collaborations, crowd-sourced solutions to problems. If done well, the global e-learning ecosystem becomes self-sustaining in the sense that each learner can eventually become a teacher or contributor, feeding back into the network.

Monitoring and Sustaining Ecosystem Health with AI and Analytics

Having theorized what constitutes health at various scales, the next question is: how do we practically monitor and maintain these complex ecosystems? This is where emerging technologies – particularly AI, learning analytics, and adaptive platforms – come into play. In a way, these tools can act as the caretakers or even the automatic regulators of an e-learning ecosystem, much like homeostasis in a living organism. But leveraging them wisely requires asking the right questions and being mindful of their limitations.

Modern learning analytics (LA) systems are designed to collect and analyze data from learning environments to help us understand and optimize them. LA is broadly defined as the “measurement, collection, analysis and reporting of data about learners and their contexts, for purposes of understanding and optimising learning and the environments in which it occurs” [9]. In the context of our ecosystem analogy, analytics can be seen as the diagnostic instruments and sensors that keep track of the system’s vital signs. For example, at the micro-level, an LMS dashboard might display which quiz questions were most frequently missed – a blood-pressure cuff detecting points of stress. At the meso-level, social learning analytics might map the network of interactions in a course, revealing whether knowledge is circulating or if there are bottlenecks. At the macro-level, analytics can highlight trends in enrollment, completion, or even alumni career paths, indicating the long-term viability of programs.

Increasingly, artificial intelligence is being layered on top of these analytics to not just report data, but to act on it. Adaptive learning engines can personalize content delivery in real-time: if a student is breezing through a topic, they get harder questions; if they’re struggling, the AI provides hints or supplemental material. We might imagine AI as a sort of autonomic nervous system for the learning ecosystem, making minute adjustments continuously to keep everything in balance. For instance, some generative AI-driven tools can summarize learning analytics findings and alert instructors to unusual patterns. According to a recent EDUCAUSE report, generative AI tools could soon report learning analytics findings in real time, allowing faculty to make data-informed decisions and interventions on the fly [6]. This suggests a future where an AI might say, “Section 2 of your course is causing a 30% drop in engagement; here are three suggested tweaks to try right now,” effectively becoming a co-instructor focused on ecosystem health.

While these possibilities are exciting, we must approach them with critical questions in mind (and MOSAIC’s practical lens at hand). Here are a few emergent questions to consider:

• Can AI-driven analytics serve as an early warning system for ecosystem health? For example, could an algorithm detect when the pedagogical metabolism of a course is slowing down – say, fewer logins or forum posts this week – and alert faculty to intervene before students start dropping out?
  
• Could adaptive platforms function as an educational immune response? If a subset of students in a program shows signs of low digital cognitive resilience, the system might automatically offer resilience-building resources: study strategy modules, motivational messages, or connect them with mentors. In essence, the technology could respond to stress in the ecosystem by shoring up support around vulnerable learners.
  
• What new metrics might we develop for holistic ecosystem monitoring? Traditional metrics (grades, completion rates) only tell part of the story. To truly gauge adaptive curricular entanglement, for instance, we might track how often students make cross-references between courses or apply a concept from one context in another. Could AI parse assignment submissions or discussion posts to find evidence of interdisciplinary thinking or real-world application? At the global scale, we might envision a “learning climate index” that measures the openness and connectivity of learning across institutions.
  
• How do we balance automation with human judgment in sustaining the ecosystem? This is perhaps the most important question. AI can crunch vast data and even make recommendations, but determining the pedagogical significance of those insights is a task for educators. An analytics system might show low activity in a forum, but only a teacher knows it’s because students are busy doing fieldwork that week. We should leverage AI to augment the visibility of the ecosystem’s state (acting as a microscope or telescope, if you will), while relying on skilled teachers and administrators to interpret and act on the information in context. The MOSAIC framework, with its blend of academic rigor and practical relevance [1], reminds us that technology is in service of pedagogy, not the other way around.

Ultimately, maintaining the health of e-learning ecosystems at scale will likely be an iterative, collaborative process. Just as ecosystems in nature benefit from biodiversity, our learning ecosystems benefit from a diversity of tools and approaches – human mentors, analytics dashboards, adaptive courseware, peer networks, and more. AI and analytics can provide unprecedented support in monitoring and nurturing these environments, but we must continually ask how their use aligns with our educational values and objectives. Are we fostering genuine understanding and growth, or just optimizing for easily quantifiable proxies? Keeping the learner at the center of the ecosystem – as MOSAIC does by design [1] – is key to ensuring that all this tech-driven monitoring actually translates into meaningful, sustainable learning.

Conclusion: Toward Thriving Online Learning Ecosystems

Thinking of online education as a living ecosystem – from the tiniest lesson to the global network of learners – allows us to apply rich metaphors and systems thinking to its design and evaluation. The MOSAIC framework offers a scaffold for this approach, ensuring that our focus on modularity, outcomes, stackability, adaptivity, and integration is maintained at every scale. A thriving e-learning ecosystem is one in which pedagogical metabolism is high (active engagement and feedback loops abound), digital cognitive resilience is strong (learners and the system can handle challenges), and curricular entanglement is adaptive and meaningful (learning experiences connect and respond to needs). It is an ecosystem that grows and improves with each cohort – feeding forward insights, embracing new technologies thoughtfully, and expanding access to those who need it.

As we innovate in online graduate education and beyond, we should remember that ecosystems are delicate. They require careful cultivation, continuous assessment, and sometimes tough interventions to remove unhealthy elements or to recover from disruption. By raising theoretical questions and framing practical monitoring strategies, we set the stage for more resilient and sustainable learning environments. In the spirit of conceptual experimentation, we have imagined new terminologies to guide our thinking. But these ideas must ultimately translate into action: new program designs, experimental uses of AI for student support, policies for credit transfer and micro-credentialing, and research that rigorously evaluates what works.

The conversation is just beginning. We invite educators, instructional designers, administrators, and researchers to carry it forward: How might your e-learning ecosystem – whether a single classroom or a global platform – be better understood as a living system? What signs of flourishing or faltering do you observe, and how might concepts like metabolism, resilience, and entanglement apply? By sharing case studies and data across institutions, we can develop a collective knowledge base on maintaining the well-being of online learning at scale. In doing so, we’ll help ensure that the rapid growth of digital education leads not to burnout or collapse, but to gardens of learning that are continuously blooming, rich with diversity, and capable of sustaining learners for the long term.

---

References

1. Cope, B., & Kalantzis, M. (2016). e-Learning Ecologies: Principles for New Learning and Assessment. Routledge.
2. Kalantzis, M., & Cope, B. (2020). After the COVID-19 crisis: Why higher education may (and perhaps should) never be the same. ACCESS: Contemporary Issues in Education, 40(1), 51–55.
3. Bevan, B. (2016). STEM learning ecologies: Relevant, responsive, and connected. Connected Science Learning, 1(1).
4. Digital Promise. (2023). The Role of Micro-credentials in the Credential Ecosystem.
5. UNESCO Institute for Lifelong Learning. (2023). Learning ecosystems (web page statement).
6. EDUCAUSE. (2023). 2023 EDUCAUSE Horizon Report: Teaching and Learning Edition. EDUCAUSE Press.
7. Hickey, H. (2022). Bringing the Field to Students with “Virtual Field Geology”. University of Washington News.
8. Fan, S., et al. (2024). Supporting engagement and retention of online and blended-learning students: A qualitative study from an Australian University. The Australian Educational Researcher, 51(1), 403–421.
9. Viberg, O., Hatakka, M., Bälter, O., & Mavroudi, A. (2024). Closing the loop by expanding the scope: using learning analytics within a pragmatic adaptive engagement with complex learning environments. Frontiers in Education.
10. Preply. (2021). Enrollment for online courses is skyrocketing: 2021 Coursera Impact Report. Retrieved from Preply blog: https://preply.com/en/blog/online-learning-statistics/.
11. Pettijohn, J. C. (2025). Cultivating e-learning ecosystems: Designing digital ecologies for environmental geology graduate education. Online Graduate Innovation. Retrieved from https://publish.illinois.edu/online-grad-innovation/cultivating-e-learning-ecosystems-designing-digital-ecologies-for-environmental-geology-graduate-education/. Note: The MOSAIC framework model used in this blog post has been adapted and updated based on this original publication.
12. eLearning Industry. (2024). Area9 Learning Platform Review: AI-powered adaptive learning system for personalized education. Retrieved from https://elearningindustry.com/directory/elearning-software/area9-learning-platform.
13. University of Washington. (2022). Bringing the field to students with “Virtual Field Geology.” University of Washington News. Retrieved from https://www.washington.edu/news/2022/12/08/virtual-field-geology/.
14. Government of Northwest Territories. (2024). How is aquatic ecosystem health measured? NWT Water Stewardship. Retrieved from https://www.gov.nt.ca/en/services/water-stewardship.